计算机技术与发展2012，Vol.22Issue(7)：20-23,4.
基于改进的混合高斯模型的运动目标提取
Moving Object Extraction Based on Improved Gaussian Mixture Model
摘要
Abstract
Background extraction is a key step for image and video processing technology. In this paper,the moving object extraction in the static background is studied, hi recent years the Gaussian mixture algorithm received extensive attention. The traditional algorithm model each pixel a fixed number of components, which is not optimal in term of detection and computational time. And the algorithm is sensitive to the adjustment of the teaming rate. In this paper, improved adaptive algorithm is put forward for moving object extraction. The major improvement is the number of mixture Gaussian components and the discriminant criterion. The experiment results show that the improved algorithm is better man traditional algorithm in both adaptability and computing speed.关键词
背景建模/运动目标提取/混合高斯模型/序列图像分析Key words
background modeling/moving object extraction/mixture Gaussian model/sequential image analysis分类
信息技术与安全科学引用本文复制引用
杨宁,杨敏..基于改进的混合高斯模型的运动目标提取[J].计算机技术与发展,2012,22(7):20-23,4.基金项目
南京邮电大学攀登计划(NY208050) （NY208050）
